Lifetime distribution of information diffusion on simultaneously growing networks

This paper studies information diffusion modeled by the SIS epidemic model on several classes of growing networks. It is shown through a thorough simulation study that there is a fundamental difference in the behavior of epidemic processes on growing temporal networks in comparison with the same processes on static networks. An infection under the SIS model tends to persist longer on a growing network than on a static network, and further, the empirical distribution of the lifetime of an infection on growing networks has a considerably heavier tail. In fact, evidence is provided that, under certain combinations of model parameters, it may be possible for an infection to survive for an infinite amount of time. These results are observed for the SIS model acting on growing temporal networks generated via the preferential attachment model and a uniform attachment model, as well as a real-world, time-stamped citation network.